Podcast Analytics: A Complete Guide to Measuring Your Show

Podcast analytics has a reputation problem. Half the industry treats download counts as gospel. The other half has decided measurement is impossible and stopped looking. Both camps are wrong, and both waste budget because of it.

The truth sits somewhere in the middle. Podcast data is messier than web analytics, scattered across more dashboards than anyone would like, and full of metrics that sound identical but mean very different things. Once you know what to measure, where the numbers actually live and which ones to distrust, you can run your show on evidence rather than vibes.

This guide covers the lot: the metrics worth your attention, the platforms that hold them, the reasons headline numbers mislead, and how to connect a show to real business outcomes.

What Podcast Analytics Can and Cannot Tell You

Start with honest expectations. Podcasting has no single source of truth. One episode gets consumed on Apple Podcasts, Spotify, YouTube and a long tail of smaller apps, and every platform reports its own numbers in its own dashboard, using its own definitions. There is no equivalent of Google Analytics quietly watching the whole thing.

Here is what you can know with reasonable confidence:

• How many times each episode was downloaded or streamed
• How long listeners stayed, and where they dropped off
• Whether your audience is growing, shrinking or flat
• Broad audience demographics on Spotify and YouTube
• With some setup, which listeners go on to become leads and customers

And here is what you cannot know: the name of every listener, a complete cross-platform picture in a single login (without doing some assembly work), or a perfect attribution trail from episode to invoice. Anyone promising those things is selling something.

That is the deal. It is a better deal than most marketing channels offer, incidentally. If you would rather someone else did the assembly work, our podcast analytics service handles measurement and reporting for B2B shows. The rest of this guide explains the principles behind it.

The Metrics Worth Measuring

Downloads and streams

The industry's default currency. A download is recorded when an app requests your audio file from your hosting platform. It proves the file was fetched. It does not prove anyone pressed play, which we will come back to.

Downloads are still useful. Tracked consistently, they show trend: whether new episodes reach a bigger audience than the ones before them, and whether your back catalogue keeps earning attention. Just treat them as a measure of reach, not attention.

Consumption and retention

The metric that separates serious shows from vanity projects. Apple, Spotify and YouTube all report how much of each episode people actually consume, minute by minute. A show with 300 listeners who finish every episode is a healthier asset than one with 3,000 who leave after ninety seconds.

Retention charts also tell you exactly where to improve. A cliff two minutes in means your intro is too long. A slow, steady decline is normal. A spike means a moment worth clipping for LinkedIn.

Followers and subscribers

Followers on Apple and Spotify, subscribers on YouTube. This is your compounding asset: the group that receives every new episode by default. Follower growth is a better health measure than raw downloads because it is harder to inflate and it predicts future performance.

Watch the ratio, too. If downloads climb but followers stay flat, you are renting attention from guest audiences and promotion rather than earning a returning audience of your own.

Engagement beyond the play button

Comments on YouTube, shares of clips, replies to the newsletter, a prospect mentioning the show on a sales call. None of this sits neatly in a dashboard. All of it matters, particularly for B2B shows, where fifty of the right listeners can be worth more than fifty thousand of the wrong ones.

Business metrics

Website visits from show notes, form fills that mention the podcast, deals influenced by an episode. These need deliberate setup, covered further down, but they are the numbers your finance director will actually respect.

Where Your Podcast Data Actually Lives

Four places, for most shows. Get logins for all of them and check monthly at minimum. A long tail of smaller apps (Overcast, Pocket Casts and friends) will also play your episodes, but they all fetch the files from your host, so that listening is already captured in your hosting stats.

Your hosting platform

Wherever your RSS feed lives, that platform sees every download request from every listening app. It is the only place with a cross-app view of your audio audience: total downloads, the apps people use, countries, and episode-by-episode comparisons. Treat it as your baseline record for audio.

Apple Podcasts Connect

Apple's dashboard reports followers, plays, average consumption per episode and total time listened. Because it measures behaviour inside the app rather than file requests, it knows the difference between a download and an actual listen. The consumption data is the star here: it shows how long real people genuinely stayed.

Spotify for Creators

Spotify reports streams, unique listeners, followers and episode retention, plus aggregated audience demographics such as age and gender. For many shows Spotify is the single biggest audio platform, and this level of detail only appears in Spotify's own dashboard, not in your hosting stats.

YouTube Analytics

If your show is video-first (ours are, and we would argue yours should be too), YouTube offers the deepest dataset of the lot: views, watch time, retention curves, thumbnail click-through rates, traffic sources, demographics and subscriber conversion. YouTube is also the only major platform that tells you how people discovered the show in any useful detail. It is a large part of why we produce every show as video; our video podcast production page explains the approach.

Why Platform Stats Mislead

A download is not a listen

The oldest confusion in podcasting. For years, apps downloaded new episodes automatically in the background whether or not anyone pressed play, and download counts flattered everybody. Apple tightened this behaviour with iOS 17 in 2023, cutting automatic downloads for lapsed listeners, and plenty of shows watched their numbers fall overnight whilst their real audience had not changed at all.

The lesson stands: downloads measure file requests, consumption measures attention. Report them as different things, because they are.

IAB standards help, but only so far

The IAB (Interactive Advertising Bureau) publishes technical guidelines for podcast measurement. Hosting platforms certified against them filter out bots, deduplicate repeated requests from the same listener within a day, and only count a download once a meaningful portion of the file has been fetched. If your host is IAB certified, your download counts are conservative and comparable with other certified shows, which is what advertisers want.

What certification does not do is turn a download into proof of listening, or force Apple, Spotify and YouTube to define their metrics the same way. A play on Apple, a stream on Spotify and a view on YouTube are three different thresholds measured in three different ways. Comparing them directly is comparing apples to oranges to slightly different apples.

The numbers will never reconcile

Add your hosting downloads to your Spotify streams and your YouTube views and you will double count some people and miss others entirely. Each platform samples a different slice of your audience at a different point in the journey. The fix is not to torture the figures until they agree. Pick a consistent set of metrics, track them the same way every month, and read the trend rather than the totals.

Connecting Podcast Data to Business Outcomes

This is where most company podcasts fall down. Not because the connection is impossible, but because nobody set anything up. Download numbers land in a quarterly slide, everyone nods politely, nothing changes. If the show exists to drive pipeline, hiring or authority, measure it like it does.

The good news is that B2B shows have a shorter distance to travel here than consumer ones. The audience is smaller, easier to identify and much closer to a buying decision, so even light-touch tracking produces signals you can act on.

Practical ways to make the link:

• Trackable links. Put UTM-tagged links in show notes, YouTube descriptions and clip captions so podcast traffic shows up in your web analytics as its own source.
• A dedicated landing page. Give the show its own short URL to mention on air. Visits to it are a clean signal that listeners took action.
• Ask the question. Add 'How did you hear about us?' to demo and contact forms. Self-reported attribution is imperfect, and it will still surprise you how often the podcast comes up.
• Listen to the pipeline. When deals close, ask whether the buyer knew the show. Podcast influence is usually invisible to last-click attribution and very visible in conversation.
• Count the guest effect. On B2B shows, guests are often prospects or partners. Track what happens to those relationships after the recording.

Notice that reach barely features in that list. A show like The CFO Playbook, Soldo's podcast for CFOs, is not chasing mass-market numbers. It is built for a defined audience of finance leaders and judged on whether it reaches and holds them. Reach is a media metric. Attention from the right people is a business one.

Where Insight Studio Fits

A brief and honest word on our own product. Insight Studio is Earworm's audience intelligence and analytics platform. It pulls the numbers above into shareable dashboards, layers in B2B audience data, and handles attribution when you put paid media behind clips. It exists because assembling four logins into one credible monthly report is tedious, and because the Report stage of our process (Create, Produce, Publish, Report) needs pipeline attribution, not just download charts.

What it will not do is identify individual listeners or conjure a perfect attribution trail out of thin air. No tool can, whatever the sales deck says. It makes the achievable version of podcast measurement faster and considerably easier to share with a board.

A Simple Monthly Measurement Routine

You do not need a data team. You need thirty minutes a month and the discipline to repeat it. Book a recurring slot and keep the same spreadsheet from month one, because the value is in the comparison, not the snapshot.

1. Pull downloads from your host, plays and consumption from Apple, streams and retention from Spotify, views and watch time from YouTube.
2. Log follower and subscriber counts across all three apps.
3. Find your best and worst retaining episode, and write one hypothesis about why each performed the way it did.
4. Check UTM traffic, landing page visits and any form fills or sales conversations that mention the show.
5. Decide one change for next month. One. Then actually make it.

Do this for two quarters and you will understand your show better than most networks understand theirs. And if the numbers say the show is stalling, that is a growth problem rather than a measurement problem: our podcast growth page covers that side of the job.

Let Earworm Handle the Numbers

Earworm is a B2B video podcast agency. We create, produce and distribute shows, then measure them properly through our podcast analytics service and Insight Studio, with plans from £1,500 a month and launch in 4 to 8 weeks. If you want podcast numbers your board will actually trust, book a call.